The orderly progression of the mammalian cell cycle requires the precise and regulated interplay between growth promoting and inhibiting signals. A critical period for integrating growth regulating signals occurs during early cell cycle progression where, given the appropriate conditions, cells progress through the "restriction point" and thereafter become committed to entering into S phase and the division cycle.
The transition from G1 into S phase is governed by a number of proteins with established roles in cell cycle control. One of the principal players in the control process is the retinoblastoma protein (pRb), a protein with known tumour suppressor properties and whose gene is frequently mutated in human tumour cells (1). The Rb protein is believed to influence the transition from G1 into S phase by regulating cellular targets, such as the family of E2F transcription factors (1,2,3,4). Indeed, a considerable body of evidence supports the idea that E2F transcription factors are preeminent physiological targets in growth control mediated by pRb. For example, over expression of E2F can promote entry into S phase and cell cycle arrest imposed by pRb can be overridden by co-expression of E2F proteins (1,5,6,7,8).
The pathway regulated by pRb receives signals from members of the family of cyclin dependent-kinases, known as cdks, which govern cell cycle progression by controlling the activity through phosphorylation of critical substrates, such as pRb (9,10). Of particular importance for early cell cycle control are cyclins A, D and E which together with an appropriate catalytic cdk subunit, are believed to mediate their effects on proliferation in part by phosphorylating pRb and thereby inactivating its growth regulating properties (11,12,13). Although the contribution of each cyclin/cdk kinase to the phosphorylation of pRb is not clear, cyclins of the D class, frequently as a complex with cdk4, appear to regulate early G1 progression, a process which very likely involves the control of pRb activity (12,13,14).
Loss of normal control by pRb, either through direct mutation in Rb and the action of viral oncoproteins, or indirectly by mutation in one of the genes encoding an afferent regulator, causes E2F to be uncoupled from its control mechanisms (2,15,16). The physiological advantage of such a mechanism for an aberrantly proliferating cell can be surmised when the nature of the target genes regulated by E2F is considered, since many of them are required for S phase progression, such as the genes for DHFR, thymidine synthetase and DNA polymerase .alpha., and others which play regulatory roles during the cell cycle, such as cdc2, B-myb and cyclin A (1,3).
E2F has a heterodimeric DNA binding activity which arises when an E2F and a DP family member interact. Five E2F (17, 18, 19, 20, 21, 22, 23, 24, 25) and three DP (26, 27, 28) family members have been characterised which can interact combinatorially to generate an array of sequence specific heterodimers. Although the E2F component of the heterodimer is variable, the DP component appears to be less so with DP-1 being frequently present in many different cell (26,31,32) Binding of pRb (or pRb-related proteins) to the E2F/DP heterodimer impedes trans activation (29,30), possibly by preventing the activation domain from interacting with the basal transcription apparatus.